You may run DIGITtally on a list of genes you're already interested in, to see how "interesting" each looks in your selected tissue set. Genelists can be built using the "Build Genelist" option from the "Other Utilities" menu
Hover over the "?" next to any data source for more information on the data inculded and how we use it.
| Standard Datasets |
|
|
| Transcriptomic Datasets |
?
FlyAtlas1 comprises 44 Affymetrix Dros2 expression arrays, each mapping the expression of 18770 transcripts. The dataset thus contains over 822800 separate datapoints.
For this study, wild-type Canton S adult flies were reared at 22C on a 12:12h light regime, on standard Drosophila diet, 1 week after adult emergence. Tissues were dissected out (from equal numbers of male and female 7-day old flies except in the case of gonads) and pooled to make at least 1500 ng mRNA, then amplified and hybridised using the Affymetrix standard protocol. For each tissue, 4 independent biological replicates were obtained. Each array thus corresponds to one biological replicate.
Note that not all tissues have a larval counterpart - only Salivary glands, Midgut, Malpighian tubule, Hindgut, Fat Body, CNS and carcass do.
As FlyAtlas1 is microarray based, readings have not only expression values for each tissue, but also a "Presence" value, which informs on whether the transcipt is detected at all according to Affymetrix's GCOS software. This can be used to infer how consistently a gene is expressed in a given tissue. By default, DIGITtally requires that at least 3/4 samples from a tissue display detectable expression to classify a gene as consistently expressed in that tissue.
|
|
?
As in FlyAtlas1, FlyAtlas2 wild-type Canton S adult flies were reared at 22C on a 12:12h light regime, on standard Drosophila diet, 1 week after adult emergence. Tissues were dissected out (from equal numbers of male and female 7-day old flies except in the case of gonads) and pooled to make at least 1500 ng mRNA.
In FlyAtlas2, the use of RNA-Seq rather than microarray hybridisation removes any ambiguity in the identification of genes and allows analysis of specific RNA transcripts. Approximately 17,500 genes and 34,500 transcripts are represented.
MOST tissues in FlyAtlas2 have three RNAseq replicates within the dataset. However, the following tissues have only two replicates, so please treat these with care: Eye, Thoracicoabdominal ganglion, Crop, Salivary gland and Female Rectal Pad.
By default, FlyAtlas2 uses an FPKM value of 2 as a cutoff for background-level expression. As such, transcripts with an FPKM under 2 are taken as not expressed in the analysed tissue by DIGITtally, and Whole Fly expressions for genes have a minimum value of 2 for the purposes of enrichment analysis.
|
|
?
The result of a huge multi-centre effort, FlyCellAtlas is an enormous dataset of single-nucleus RNA sequencing data from over 580000 nuclei. These nuclei are annotated at both the tissue level (see above) and at the cell type level (over 250 unique cell types represented).
RNA samples were sequenced using either Smart-seq2 or 10x Genomics sequencing technologies - the latter dataset is utilised for DIGITtally as a "Whole Body" dataset is available. Tissues were broadly isolated by dissection except in cases where this was not possible due to localisation throughout the body - fat body, oenocyte, and trachea. In these cases, tissue-specific fluorescence was utilised to enable FACS-based nuclear isolation.
DIGITtally utilises the "Relaxed" FlyCellAtlas dataset due to the reduced stringency on mitochondrial content within analysed cells, meaning cells with high mitochondrial content (such as those within the digestive tract and malpigian tubules) are not excluded. Still, this dataset contains data on cells which have a < 15% mitochondrial content. Nuclei were also removed if they were found to be "poor quality" - ie had fewer than 10,000 uniquely mapped reads.
For our analyses, we utilise the Unique Molecular Identifier-based read counts found in the FlyCellAtlas Whole Body dataset. Following the stategy outlined by the authors of FlyCellAtlas, we use define a cell as expressing a given gene if at least one Unique Molecular Identifier corresponds to the gene.
|
|
?
MozAtlas comprises 60 Affymetrix Anopheles expression arrays performed on 5-day-old G3 laboratory strain mosquitoes, mapping the expression of ~16,000 probes - corresponding to almost 80% of known Anopheles genes. For each tissue and gender, an equal quantity of RNA was pooled from three time points (24, 48, 72hrs). In females samples were taken after the blood-meal. For each of four biological replicates, tissues were pooled from a minimum of 10 mosquitoes dissected at each time point
As MozAtlas is microarray based, readings have not only expression values for each tissue, but also a "Presence" value, which informs on whether the transcipt is detected at all according to Affymetrix's GCOS software. This can be used to infer how consistently a gene is expressed in a given tissue. By default, DIGITtally requires that at least 3/4 samples from a tissue display detectable expression to classify a gene as consistently expressed in that tissue.
|
|
?
Mosquitos of the A. gambiae Keele strain were at 26 °C and 80% humidity with a 12 h day: night cycle. Adult mosquitoes were given access to 5% glucose in 0.05% para-aminobenzoic acid water, and larvae fed with ground TetraMin tropical fish food (Tetra, USA). Adult mosquitoes were offered a human blood-meal through an artificial membrane once per week.
Malpighian tubules from approximately 400 uninfected An. gambiae larvae, sugar-fed adult males, sugar-fed adult females, and females 3 h after a blood feed were dissected. RNA was isolated, and used for a Plasmodium/Anopheles Genome array containing probe sets to approximately 14,900 An. gambiae transcripts
As MozTubules is microarray based, readings have not only expression values for each tissue, but also a "Presence" value, which informs on whether the transcipt is detected at all according to Affymetrix's GCOS software. This can be used to infer how consistently a gene is expressed in a given tissue. By default, DIGITtally requires that at least 3/4 samples from a tissue display detectable expression to classify a gene as consistently expressed in that tissue.
|
|
?
Aegypti-Atlas we uses 5- to 12-day-old female Ae. aegypti mosquitos. Larvae were reared at a density of 200 larvae per 1 liter tray, recieving 720 mg of fish food. Adults were maintained in humidified chambers (RH 75% ± 5%) at 29 °C on a diet of 10% sucrose ad libitum. Females were reared and maintained in approximately equal proportion with males for a minimum of 5 days prior to blood-feeding and/or dissection and were presumed to be mated.
Mosquitos were dissected, then RNA was harvested for RNA-seq analysis using the Illumina Nextseq 500 platform. Three biological replicates were prepared for each tissue.
To keep analysis consistent with other RNA-seq datasets, raw read counts from the Aegypti-Atlas datasets were converted to FPKMs prior to analysis. By default, DIGITtally uses an FPKM value of 2 as a cutoff for background-level expression. As such, transcripts with an FPKM under 2 are taken as not expressed in the analysed tissue by DIGITtally, and Whole Body expressions for genes have a minimum value of 2 for the purposes of enrichment analysis.
|
|
?
The SilkDB RNA-seq dataset is robust, with 253 RNA-seq datasets gathered from different tissues and stages of silkworm. RNA-seq analysis used the Illumina NovaSeq 6000, followed by HISAT2 alignment and Ballgown expression analysis. FPKM values from these analyses are used within DIGITtally.
As SilkDB lacks a true "Whole Body" gene expression reading, DIGITtally generates an artificial whole body expression level by averaging expression across all tissues present in each silkworm. This may not be truly representative of the whole body expression of a given gene due to the differences in relative contributions between different tissues at differing life stages, but should provide a reasonable estimate for enrichment calculations.
Please note that not all tissue types have data associated with each silkworm stage. DIGITtally automatically accounts for this by generating working tissue of interest lists for each selected silkworm type, including only those tissues of interest which have data associated. Final scores for Enrichment, Specificity and Expression are thus split in two, with a half-point for passing the threshold in all applicable tissues in ONE silkworm type, and a second for passing the threshold in ALL selected silkworm types.
By default, DIGITtally uses an FPKM value of 2 as a cutoff for background-level expression. As such, transcripts with an FPKM under 2 are taken as not expressed in the analysed tissue by DIGITtally, and Whole Body expressions for genes have a minimum value of 2 for the purposes of enrichment analysis.
|
|
|
| Other Datasets |
?
FlyBase is the major repository for Drosophila melanogaster information, with individual pages for each gene within the fly containing a wealth of data from across the field. Indeed, many of the data sources used to build DIGITtally are key parts of the FlyBase gene summary.
As FlyBase contains information from >100000 INDIVIDUAL research papers, along with thousands of notes from direct personal communications, it is the ideal database to use for a directed literature search for for associations between a set of genes and a given set of tissues of interest.
FlyBase also contains a comprehensive list of the mutant phenotypes associated with alleles of each Drosophila gene. As such, we are able to score genes of interest based on whether they are known to affect specific tissue functions.
|
|
?
Many tools exist to find orthologs of a gene between different species - particularly those species which have well annotated genomes. However, these tools rarely provide the same results, making it difficult for researchers to know which to use. DIOPT seeks to rectify this issue.
DIOPT integrates 17 different orthology finder algorithms for 10 different model organisms. For each gene, the number of orthology algorithms which return a given ortholog are taken as the "DIOPT score", with a higher score representing a more reliable ortholog. By default, DIGITtally only considers orthologs with a "Moderate" or better DIOPT score (3 or greater) as reliably orthologous.
|
|
|
| User-Uploaded Data |
|
|
| How to upload your data |
Change Data Source Weighting
These options change the relative weighting of scores from each data source - ie, if you set FlyAtlas1 to 1 and Flyatlas2 to 2, scores from FlyAtlas2 will contribute twice as much to your DIGITtally.
Please note, if you set a weight to 0, that data source will no longer be used.
